SECURITY CLASSI~CATION OF THI PAGE (We Dat ACCESSeON NO

REPORT DOCUMENTATION PAGE '4o""REAFISRCINBEFORE COMPLETING FORM

I. RPOR NUMER OVT CCESIONNO.3. RECIPIENT'S CATALOG NUMBER

Technical Report No. 5 -P - //5 3 S4. TITLE (and Subtitle) 5. TYPE OF REPORT & PERIOD COVERED

Structure and Adsorption Characteristics of Silane Technical ReportCoupling Agents on Silica and E-glass Fiber; S. PERFORMING ORG. REPORT NUMBER

Dependence on pH7. AUTHOR() S. CONTRACT OR GRANT NUMUER(s)

S. Naviroj, S. Culler, J.L. Koenig, and H. Ishida N00014-80C-0533

9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT, PROJECT. TASK

kn ~Department of Macromolecular Science AE OKUI UBR

vm4 Case Western Reserve University NR 356-7391Cleveland, Ohio 4410612REOTDE

rmq 11.CONTROLLING OFFICE NAME AND ADDRESS12REOTDE

Office of Naval Research June 2. 1982ot 800 N. Quincy Street 13. NUMBER OFPAGES

Arlington, VA 22217 3414 OITORING AGENCY NAME & AODRESS(ll different from Controlling Ollice*) iS. SECURITY CLASS. (of this report)

Unclassified

15a. DECL ASSI FICATION/ DOWNGRADINGSCHEDULE

16. DISTRIBUTION STATEMENT (of (hie Report) D IApproved for public release; distribution unlimited ELECTE

JUN 9

17. DISTRIS3UTION ST 4ENT ('of ''abstract entered in Block 20, It different from, Report)

III. SUPPLEMENTARY FE S

Prepared for publication in journal

19. KEY WORDS (Continue on revesee side it necessary and Identify by block nuotbor)

Silane Coupling Agents Adsorption CharacteristicpH Effects SilicaE-glass Fiber Fourier Transform Infrared Spectroscopy

20. AS ACT (Continue an reverse aid* if necoeery, and identify by block number)

Fourier transform infrared spectroscopy was used to study the adsorptionof silane coupling agents on silica and E-glass fiber surfaces. The structures

Liiof the adsorbed species of(-'aminopropyltriethoxysilane (y-APS) coupling agentunder various preparation conditions are postulated on an interpretatiton of theNH deformation and SiGH stretching modes. The amount f APS deposited on

contrary, the amounts of adsorption of(- methacryloxyprop trimethoxysilane

DD~ DORM 1473 Please see r v rse sideSECUIRITY CLASSi P 9N" PAGE (When Data Entered)

S9CMI? CLAS31FICATION OF TWOS PAGE(I~imi Data Enteu.el)

20. ABSTRACT (continued)

(y-MPS) and vinyltriethoxysilane (VS) coupling agents are independent ofpH. The number of adsorbed molecules of y-'MPS and VS is comparable to that

ol1'-AS only at its acidic pH. The unique adsorption of y-APS on the..ilc surface is partly due to the structure of the y-APS itself in the

/"-'reating solution. When E-glass fiber is used as the substrate, the amountof y-APS adsorption is affected by pH but less than observed with silicapowder. The structure of y-APS is affected by the presence of carbon di-oxide when the pH range of the silane solution is between 6 and 12.

Accesaion For

NTIS GRA&IDTTC TABQUnpurinounoed

Availability CodesAvail and/or

Dist Special

SECURITY CLASSIFICAT ION OF TMflS PA~Gru1~.f Dae Enteted)

Structure and Adsorption Characteristics of Silane Coupling Agents

on Silica and E-glass Fiber; Dependence on pH

by

S. Naviroj, S. Culler, J.L. Koenig, and H. Ishida

Department of Macromolecular Science

Case Western Reserve University

Cleveland, Ohio 44106

ABSTRACT

Fourier transform infrared spectroscopy was used to study

the adsorption of silane coupling agents on silica and E-glass

fiber surfaces. The structures of the adsorbed species of y-

aminopropyltriethoxysilane (y-APS) coupling agent under various

preparation conditions are postulated based on an interpretation

of the NH deformation and SiOH stretching modes. The amount of

y-APS deposited on high-surface-area silica depends on the pH

of the treating solution. On the contrary, the amounts of ad-

sorption of y-methacryloxypropyltrimethoxysilane (y-MPS) and

vinyltriethoxysilane (VS) coupling agents are independent of

pH. The number of adsorbed molecules of y-MPS and VS is com-

parable to that of y-APS only at its acidic pH. The unique

adsorption of y-APS on the silica surface is partly due to the

structure of the y-APS itself in the treating solution. When

E-glass fiber is used as the substrate, the amount of y-APS

adsorption is affected by pH but less than observed with silica

powder. The structure of y-APS is affected by the presence of

carbon dioxide when the pH range of the silane solution is between

6 and 12.

*4. . . . . --.- -

2

INTRODUCTION

The surface treatment of glass fibers with silane coupling agents

has enhanced the performance of the fiber reinforced plastics [1,2].

The interaction of the silane coupling agent with the glass surface may

vary depending on the composition of the glass surface and the type of

silane coupling agent. Many theories have been proposed to explain the

function of silane coupling agents treated on glass fibers [3-10]. There

is some evidence to support each of these theories but none of the theo-

ries is conclusive enough to explain all the data. The mechanism by which

silane coupling agents function in composites remains undetermined.

When glass fibers are treated with coupling agent, the mechanical

performance of composite materials is significantly improved under humid

environments. The resistance to water may be explained by either the

formation of hydrogen bonds to the surface, electrostatic bonding by the

attraction of oppositely charged ions, or covalent bonding to the sub-

strate surface. Surfaces of metal oxides and silicates in equilibrium

with atmospheric moisture consist of surface hydroxyl groups. The

surface hydroxyl and oxide groups are important for the formation of

each type of bonding.

The electrostatic surface potential (1-potential) depends on the

concentration of H+ ion. The isoelectric point of a surface is deter-

mined by the pH at which the potential of the surface is zero. Vander-

bilt [11] indicated that different mineral surfaces have different

interactions with the silane coupling agent. The pH of the treating

solution of the glass fibers is an essential factor in controlling the'II

population of the acid sites. The chemical stability of silanol groups

is highly pH dependent. Thus, the structure of the coupling agent in

the treating solution is also strongly influenced.

Determination of the mode of silane adsorption on the glass fiber

or mineral filler surfaces may be important in understanding the re-

inforcement mechanism of FRP. Plueddemann [12] studied the catalytic

and electrostatic effects in bonding through silanes. He suggested

that electrostatic forces at the mineral surface may be important in

determining the orientation of silane molecules applied from an aqueous

solution. Cationic organofunctional silanes are sensitive to the pH of

the treating solution. The molecular orientation of cationic silane

coupling agents is affected more than the nonionic coupling agents. At

optimum condition of the treating solution, the cationic coupling agent

may be more effective than the nonionic coupling agent due to the more

favorable molecular orientation. The surface isoelectric point (IEP)

also has an effect on the molecular orientation [12]. The deposited

silane could be right side up (silanol on the mineral surface) or

upside down (the functional group on the mineral surface) depending

on both the IEP and pH of the solution.

Many researchers have studied the adsorption of silane on glass

fibers and silicate surfaces. Ishida et at [13] utilized FT-IR to

study the nature of vinyltriethoxysilane on fumed silica. They reported

that the silane chemically reacted with the silica surface. Gent et

at (14] showed that model silane compounds form chemical bonds with

silica in the presence of n-propylamine probably via a base-catalyzed

interchange reaction involving silanol groups on the silica surface.

4

Koenig et at [15] used Raman and infrared spectroscopy to study the

reactions of silane with silica and glass surfaces. Different types

of silanol groups of silica surface were detected. Kaas et at [16]

studied the interaction of silane coupling agents with silica surfaces

and found that the major forces holding the silane to the silica sur-

face after application from dilute solution are primary chemical bonds.

All these studies support the existence of chemical bonds at the glass/

silane interface.

Both the optimum mechanical performance of composites and the

amount of adsorbed silane on the surface of glass fibers depend on

the concentration of the treating solution. It is well known that

the structure of an aminosilane coupling agent is pH dependent and

also affected by drying conditions [17]. With different structures

that are influenced by the pH of the solution, the mode of adsorption

of this coupling agent may vary and the amount of adsorbed silane may

be influenced. Thus, the mechanical properties would be affected.

A high-surface-area silica was chosen to model the E-glass fiber

since silica is the major component in E-glass fiber. Study of the

interaction at the silica-coupling agent interface is facilitated by

the use of high-surface-area silica. Fourier transform infrared spec-

troscopy (FT-IR) was chosen to obtain information about the molecular

structure and the amount of silane coupling agent present on the sur-

face of silica. Our aim is to elucidate the complex nature of the

silane coupling agents on glass fiber surfaces and understand the

reinforcement mechanism.

Ai

t

5

EXPERIMENTAL

A) FT-IR analysis of partially cured X-APS

y-aminopropyltriethoxysilane (y-APS was purchased from Petrarch

Systems Inc. y-APS was mixed with deionized distilled water 20% by

weight solution. The silane was hydrolyzed for 40- 50 minutes. The

hydrolyzed y-APS was used at its natural pH or adjusted to the appro-

priate pH values by adding NaOH or HCI. The solution was cast onto

a silver bromide (AgBr) plate and dried in air at room temperature.

The dried y-APS was then examined by FT-IR. Special precautions were

taken for high pH samples. The solution was first allowed to be par-

tially dried on a glass slide. When the solution became viscous, it

was then quickly transferred to the AgBr plate for complete drying.

This procedure prevented the possible decomposition of the AgBr plate.

B) Titration

Two ml of y-APS were hydrolyzed in 20 ml of deionized distilled

water. The solution was hydrolyzed for 40 - 50 minutes. One ml of

solution was titrated with O.lN HC solution. After each addition of

HCI, the pH of the solution was allowed to stand for ten minutes to

reach equilibrium. The amount of HC added was recorded along with

the pH change. The pH of the solution was measured with a Beckman 3500

digital pH meter.

C) Silica wafers

Silica powder used in the experiment was obtained from Degussa Inc.

The silica powder has a specific surface area of 130 m2/g. The powderI

t.77

6

was made into wafers by compressing the powder between aluminum foils

under a pressure of 15,000 psi. Thin wafers with the thickness of approxi-

mately 0.1 mm were obtained. The wafers were then heated at 350C for

three days in an oven before being treated with coupling agents.

One percent by weight y-APS solutions were prepared. The solution

was adjusted to the appropriate pH by using either NaOH or HCI. Thin

wafers were treated with hydrolyzed y-APS in a Petri dish. The wafer

of approximately 3- 4 cm2 was immersed in 50 ml of solution for 15 min.

After the allocated time, the solution was decanted and the wafer was

air dried overnight at room temperature.

Infrared spectra of the samples were obtained with a Digilab FTS-14

spectrometer at a resolution of 2 cm-1 . The number of scans was 200

for the sample and 100 for the reference. The wavenumber was calibrated

with a helium neon laser with a reproducibility of 0.01 cm-I. The

spectra obtained were stored on a magnetic tape for further examination.

The same procedure was followed for other samples of silica wafers

but the coupling agents used were y-methacryloxypropyltrimethoxysilane

or vinyltriethoxysilane.

D) Calibration curve of y-APS

y-APS coupling agent was smeared on a glass slide to be hydrolyzed

by adsorbing the moisture from air. After exposure in air for ten days,

the dried coupling agent was scraped from the glass slide. The dried

powder was placed in an aluminum pan and the weight was measured by

a Perkin-Elmer microbalance. The known weight of coupling agent was

mixed with approximately 200 mg of KBr powder and compressed into a

7

pellet for infrared examination. Extra precaution was taken to ensure

that all of the coupling agent powder was transferred from the mortar

into the die of the pellet maker.

E) Calibration curve for silica

A high-surface-area silica powder was pressed into a thin wafer,

and subsequently heated at 350 0C for three days. The wafer was broken

into an appropriate size and was placed between the two Kr plates in

order to hold the wafer in place during infrared analysis. The wafer

was removed from the sample holder and weighed with the electrobalance.

The weight of silica wafer versus the absorbance at 1970 cm-1 was plotted.

F) E-glass fibers treated with 1% y-APS at various pH

E-glass mats (Crane Glas, grade 50-01) were heat cleaned in air

at 500C for 24 hours prior to use. The pH-adjusted, 1% by weight y-

APS solutions were used. Glass mats were immersed into the silane solu-

tion for one minute and the excess solution was filtered off by using

a house vacuum in order to improve the reproducibility of the experiment.

The treated mat was dried in air at room temperature for one day. About

1 mg of glass mat was ground with approximately 200 mg of KBr powder for

FT-IR analysis.

RESULTS

After hydrolyzing y-APS, the pH of the solution was adjusted in

order to study the effect of pH on the structure of the y-APS. Figure

1 shows the spectra of the samples at various pH values. For the spectra

in the pH range of 2 to 7, the bands at 1610 and 1505 cm-I were assigned

8

to the NH3 deformation modes [17]. These assignments can be supported by

the presence of the broad bands at 2800 and 2400 cm-I (not shown in fig.

1). As the pH value of the solution is raised to about 8, changes begin

to occur. We will first focus our attention on the bands around 1600- 1500

cm-l,and later the band around 930 cm-1 .

A new peak begins to appear at 1570 cm-1 from pH 8 to 10. At pH 10,

the peaks are clearly defined at 1575 and 1505 cn-l. These two bands are

+assigned to the NH3 deformation modes of aminebicarbonate salt [17]. The

presence of the bicarbonate ion can be observed in the pH range of 7 and

10.8. The band at 1332 cm- was used to monitor the amount of the bi-

carbonate group. This band is due primarily to the OC02 mode. The

intensity of this band (1332 cm-1) increases noticeably as the pH value

increases from 8.5 to 10.8. The increase in intensity is due to the in-

crease in concentration of the bicarbonate salt. As we followed the changes

further to a higher pH value, we observed that the bands at 1575 and 1485

cm disappeared, leaving a prominent peak at 1592 cm-r, which was assigned

to the NH2 bending mode. By following the structure changes due to the pH

of the solution, we observed that the structure of the partially cured y-

APS hydrolyzate changed from NH3 to NH3(HC03) and to NH2.

The peak around 930 cm-I1 which is assigned to the SiO stretching mode

of the silanol [18] can also be followed. We monitor the frequency of the

SiOH stretching mode and plot the peak position versus the pH of the solu-

tion as shown in fig. 2. The peak position of the SiOH mode remains constant

at 902 cm- at pH 1 to about pH 8. In this region of pH, the silane is in

the form of weakly hydrogen bonded SiOH and the amine group is in the form

+of NH3 . This result is in agreement with the result described earlier.

As the pH of the solution becomes higher than 8, the frequency of the SiOH

9

peak shifts upward rapidly (fig. 2). In this region the amine group, which

+was in the form of NH3 at low pH, changes its structure to form a hydrogen

bond with the SiOH, which results in the upward shifting of the SiOH peak.

Figure 3 shows the proposed structures of y-APS in a partially cured state

when prepared from solution at various pH's. The frequency of the amine

deformation and SiO stretching modes were used as a guide to the proposed

structures. At pH= 2, the two amine bands at 1610 and 1505 cm-I are indi-

cative of the NH3 structure. The band at 902 cm-I is the weakly hydrogen

bonded SiOH stretching mode. The sample dried in air from a solution of

pH= 10.6 has amine bands at 1575 and 1488 cm-l which indicate the presence

of the NH3 structure. The SiOH peak is now shifted to 930 cm-1 from 902

cm-1. This shifting is due to the existence of stronger hydrogen bonding

of the SiOH group. With the presence of bicarbonate salt, the structures

as shown in fig. 3 are proposed.

Titration was performed on the hydrolyzed y-APS. The pH of the natural

solution at 1% by weight of Y-APS has the value of 10.6. The pKa value of

ethyl amine is 10.64. [19].

The pressed silica powder does not lose its specific surface area at

this pressure [7]. Figure 4 shows the spectra of the silica wafer treated

with y-APS at various pH's. The bands at 2930 and 2860 cm-r, which are due

to the CH2 stretching modes of the propyl chain, increase in intensity as

the pH of the treating solution changes from 2 to 10.6. This indicates that

more y-APS is adsorbed at pH 10.6 than at pH 2. The silane uptake is low-

ered at the pH higher than 10.6.

In order to determine the number of y-APS molecules adsorbed on the

silica wafer versus the pH of the treating solution, the calibration curve,

10

the integrated absorbance of y-APS versus the weight of y-APS, must be

established. Dried y-APS was carefully weighted and pressed into a KBr

pellet. The spectra were taken and the area of CH2 stretching bands at

2930 and 2860 cm-1 were used to measure the amount of y-APS. The area

under the two peaks was plotted against the weight of dried y-APS as shown

in fig. 5. The slope of the line was found to be 280.7 cm-1 /mg by the

least squares method. This indicates that 1 mg of y-APS will have the

area of 280.7 cm- 1 due to the CH2 stretching bands. From the calibration

curve of silica (fig. 6), in which the absorbance of the peak at 1970

cm -1 was used to indicate the amount of silica powder, the weight of the

silica wafer treated with coupling agent can be normalized. With the

calibration curve of silica and the calibration curve of y-APS, the number

of y-APS molecules deposited per milligram of silica can be obtained.

Figure 7 shows the plot of the number of y-APS molecules per milligram

of silica versus the pH of the treating solution.

The same procedure was followed for the samples treated with y-MPS

and VS coupling agents (figs. 8 and 9). The specific absorptivity of

C - C group of unhydrolyzed y-MPS at 1720 cm- 1 was measured to be 1.924 x

107 cm/mole and was used to calculate the number of molecules present

on the silica surface. The specific absorptivity of VS coupling agent

by using the C = C stretching mode at 1602 cm -1 was 2.87 x105 cm/mole (20].

The number of y-MPS and VS molecules adsorbed per milligram of silica

powder are shown in figs. 10 and 11, respectively.

The adsorption of y-APS on E-glass fiber was determined as a function

of the pH of the treating solution (1% by weight). From digital subtrac-

tion of pure E-glass spectrum from the spectrum of the treated E-glass,

11

the amount of y-APS adsorbed on the fiber was determined and plotted in

fig. 12.

Figure 13 shows the difference spectra of untreated silica (fig.

13A) from the aminosilane-treated silica at pH 2 (fig. 13B), pH 10.5 (fig.

13C), and pH 12 (fig. 13D). The difference spectrum at pH 2 shows two

peaks at 1610 and 1507 cm 1 which are indicative of the NH3 structure of

y-APS. In the difference spectrum of pH 10.5, the peak at 1332 cm-1 is

an indication that some of the amine groups form aminebicarbonate struc-

tures. In addition, the peak at 1596 cm- 1 resembles that of the bulk

sample made from the solution of pH 12. The difference spectrum of

untreated silica from the treated silica of pH 12 solution (fig. 13D),

shows a peak at 1592 cm- 1 which is assigned to the free amine.

Figure 14 shows a difference spectrum of the y-MPS on silica. The

peak at 1705 cm- 1 was assigned to the carbonyl of the methacryl group.

DISCUSSION

The structures of Y-APS hydrolyzate in solution and adsorbed on a

glass surface are greatly dependent on the pH of the solution. It has

been shown that y-APS can be in the form of NH3C, NH3(HCO3) , free NH2,

and hydrogen bonded NH2. From the frequency shifts of the SiOH peak and

the assignment of the bands around 1600 cm- 1, the structures of y-APS

are proposed. From the IR experiments, we observe that there is a change

in the structure of partially cured y-APS above pH 8.5, which is indicated

by the frequency shift of the SiOH peak from 905 to 930 cm-1 . The influence

of the carbon dioxide on the formation of aminebicarbonate is also noticed

between pH 6 and 12. The amount of bicarbonate salt shows a maximum at

12

the natural pH of the solution. When the amount of y-APS adsorbed on

silica is examined, the maximum adsorption is also found at the natural

pH of the solution. The number of molecules adsorbed on silica from y-

APS solution at natural pH is one order of magnitude higher than the

sample treated from acidic solution. With a change of the pH from 10.6

to 2, an amine salt (NH3C) is formed. It can be seen that the amount

of adsorption is highly dependent on the structure of the amine group

itself. The high adsorption of y-APS on silica at the natural pH of the

treating solution may be explained by the difference in the interaction

of the amine group with the surface silanol of silica. At low pH, the

amine group has a strong interaction with the added acid and forms an

amine salt (- NH3CZ-). The SiO concentration of the silica surface

species decreases at low pH, whereas, the concentration of the surface

silanol increases. At natural pH the concentration of SiO on the silica

surface is higher. The positively charged amine group prefers to inter-

act with a negatively charged silica surface. However, at low pH the

tendency for the NH3CZ to interact with SiOH is lower. Therefore a

greater adsorption is observed at natural pH than in an acidic solution.

At high pH (above 11), some of the y-APS silanol groups may be in the

form of SiO which will have a tendency to repel the negatively charged

surface SiO. The adsorption of y-APS at pH 12 is lower than at natural

pH.

When y-MPS and VS were used to treat silica wafers, their adsorption

characteristics showed little iH dependency except at pH above 10. VS

adsorbs more on silica surface than y-MPS. The high adsorption of VS on

silica may be explained by the size effects. The VS molecule is smaller

13

than y-MPS, thus the silica surface can adsorb more molecules per unit

area. From the plot in figs. 12 and 13, 1.13x 1017 molecules of VS are

adsorbed per milligram of silica compared with 6.21 x 1016 molecules for

y-MPS. The VS molecules are small and can be trapped in the cavities

of the rough silica surface. Knowing the specific surface area of silica

powder (130 m 2/g), the number of molecules on the surface, and an approxi-

mate molecular size, the equivalent number of layers can be calculated.

The estimated area occupied by a VS and a y-MPS molecule are 40 12/

02molecule [13] and 50 A /molecule, respectively [21]. The calculation

shows that VS and y-MPS form 0.35 and 0.24 equivalent layers on silica

when treated with a 1% by weight solution, both of which correspond to

nearly effective monolayer coverages.

The adsorption of y-APS on E-glass mat shows a trend similar to the

adsorption on silica wafer. The maximum adsorption occurs when the treating

solution is at its natural pH. However, the absolute amounts of adsorbed

molecules at natural pH and at acidic pH do not differ as much as for the

silica wafer. This is probably due to the fact that the nearly effective

monolayer coverage on silica results in direct interaction between the

surface and the silane. Whereby, with E-glass fibers, the multilayer

adsorption shows the reduction of the surface effects; however, the ori-

entation of the silane in the first layer influences the adsorption of

the next layer and so on, thus resulting in the observation of the surface

effects on the adsorption curve but less than the silica sample.

y-APS adsorbed on silica surface may very well have more than one

structure. From the difference spectrum (fig. 13B), the peak at 1332

cm-1 is evidence that the amine bicarbonate structure is present. The

14

peak at 1596 cm-1 is also present when the sample is treated from solution

at natural pH. This band is due to the hydrogen bonded amine. Preliminary

results show that about 50% of the amine group is in the form of the bi-

carbonate. Figure 14 shows that the carbonyl group is hydrogen bonded.

In general the carbonyl peak of y-MPS appears at 1720 cm- 1 and can be

shifted as much as 1740 cd-1 upon polymerization of the C- C functional

group. The band at 1705 cm- indicates that the carbonyl is hydrogen

bonded. No direct evidence has been obtained thus far on the hydrogen

bonding of the amine group with the surface though all indirect evidence

suggests the interaction with the surface. In the case of y-MPS, the

hydrogen bonded C=O has a frequence at 1705 cm- 1. The partially cured

hydrolyzate of y-MPS contains a significant amount of SiOH which inter-

molecularly hydrogen bonds with the C = 0 group of the methacryl moiety.

For this hydrogen bonding, the C O stretching mode occurs at 1700

cm 1 [20]. This frequency shift can be considered as the evidence that

the C= 0 group is in fact interacting with the silica surface.

The structures of partially cured y-APS are proposed based on the

bands around 1600 cm-1 and the SiOH band around 930 cm-1 . The shifting

of the SiOH mode is due to hydrogen bonding. When the y-APS sample is

dried in a nitrogen atmosphere, the SiOH peak shifts to 934 cm 1 , even

higher than the sample dried in air which gives rise to the SiOH mode

at 930 cm-I. This indicates that the sample dried in nitrogen forms

stronger hydrogen bonds than the sample dried in air since stronger

hydrogen bonding increases the SiO stretching frequency of the silanol

whereas the OH stretching frequency decreases under the same circumstance.

15

It is reasonable that the sample dried in air will have less hydrogen

bonding since the bicarbonate structure is also present. At pH 12, the

amine group is mostly in the form of NH2 as indicated by a single peak at

1595 cm-1 . However, there are some residual SiOH groups indicated by the

presence of the band at 925 cm-I.

The influence of carbon dioxide on the formation of the bicarbonate

salt with the amine group is noticeable at pH 6. When y-APS is applied

at a solution pH higher than 6 onto glass surface, the bicarbonate salt

may be formed during the drying process. It is not known whether the

bicarbonate structure can alter the reactivity of the amine group with

the resin matrix. It was found that the strength of a silica-filled

polyester composite may be improved more than 100% by changing the pH

of silane application from 12 to 2 [12].

CONCLUSIONS

The structure of y-APS is highly dependent on the pH of the treating

solution. The maximum number of molecules are adsorbed when the silica

is treated with a solution at its natural pH (10.6). It is evident that

the mode of adsorption of y-APS on silica and E-glass fibers is pH de-

pendent. The amount of Y-MPS and VS adsorption, however, is independent

of the pH of the treating solution below pH 10. With 1% by weight solu-

tion, the number of molecules of VS and y-MPS on a high-surface-area silica

are 1.13x 1017 and 6.21 x 1016, respectively. Various structures were

proposed based on the amine deformation bands around 1600 cm-1 and the

SiOH stretching band around 930 cm-l. The peak of SiOH can shift from

905 cm- 1 of acidic sample to 930 cm-1 of the natural pH sample. This

shifting is due to the variation in the strength of hydrogen bonding.

7----

16

The structure of y-APS changes noticeably in the pH range of 6 to

12. At the natural pH of the treating solution, the maximum amount of

bicarbonate formation as well as the maximum amount of adsorption were

observed.

The structure of y-APS on the silica surface treated at pH-2 is

in the form of a protonated amine. At natural pH (10.6), two structures

are observed, namely hydrogen bonded NH2 and amine bicarbonate. At pH-

12, y-APS is mainly in the form of NH2 .

ACKNOWLEDGMENT

The authors gratefully acknowledge the financial support of the

Office of Naval Research under grant no. N00014-80C-0533.

17

REFERENCES

1. G.A. Park, Advances Chem. Series 6-7, 121 (1964).

2. R.J. Wong, 3. Adhesion 4(2), 171 (1972).

3. E.P. Plueddemann, "Composite Materials", Vol. 6, in Interfaces in

Polymer Matrix Composites, Academic Press, New York, 1974.

4. C.A. Kumins and J. Roteman, 3. Polym. Sci. i(lA), 527 (1963).

5. R.C. Hooper, Proc. SPI 11th Ann. Tech. Conf., Sec. 8-B (1956).

6. E.P. Plueddemann, H.A. Clark, I.E. Nelson, and K.R. Hoffman, Mod.Plast. 40, 133 (1963).

7. O.K. Johannson, F.O. Stark, G.E. Vogel, and R.M. Fleischman, J. Comp.Mater. 1, 278 (1967).

8. P.W. Erickson, A. Volpe, and E.R. Cooper, Proc. SPI 19th Ann. Tech.Conf., Sec. 21-A (1964).

9. J.0. Outwater, Proc. SPI 11th Ann. Tech. Conf., Sec. 9-B (1956).

10. W.A. Zisman, Proc. SF1 19th Ann. Tech. Conf., Sec. 21-B (1964).

11. B.M. Vanderbilt and R.E. Clayton, Ind. Eng. Chem. Prod. R. and D.____(l), 16 (1965).

12. E.P. Plueddemann and G.L. Stark, Proc. SF1 28th Ann. Tech. Conf.,Sec. 21-E (1973).

13. H. Ishida and J.L. Koenig, J. Colloid and Interface Sci. 64(3), 555

(1978).

14. A.N. Gent and E.C. Hsu, Macromolecules 7, 933 (1974).

15. J.L. Koenig and P.T.K. Shih, Materials Sci. & Eng. 20, 127 (1975).

16. R.L. Kaas and J.L. Kardos, Polym. Eng. & Sci. 11(l), 11 (1971).

*17. S. Naviroj, J.L. Koenig, and H. Ishida, Proc. SF1 37th Ann. Tech.

Conf., Sec. 2-C (1982).

*18. A. Streitwieser Jr. and C.H. Heathcock, Introduction to OrganicChemistry, MacMillan Publishing Co., Inc., New York, 1976.

19. H. Ishida, C. Chiang, and J.L. Koenig, Polymer 23, 251 (1982).

18

20. H. Ishida and J.L. Koenig, Proc. SPI 32nd Ann. Tech. Conf., Sec. 4-B(1977).

21. H. Ishida, S. Naviroj, K. Tripathy, J.J. Fitzgerald, and J.L. Koenig,J. Polym. Sci., Polym. Phys. Ed. 20, 701 (1982).

19

FIGURE CAPTIONS

Fig. 1 FT-IR spectra of partially cured y-APS. The solution at variouspH was dried on AgBr plate in room environment.

Fig. 2 The position of SiOH peak of partially cured y-APS shifts as a

function of pH of the solution.

Fig. 3 Proposed structures of partially cured y-APS as a function of pHof the solution and drying conditions. Amine deformation and SiOHvibrational modes were used to derive the structures.

Fig. 4 FT-IR spectra of (A) untreated silica wafer and (B) treated silicawith 1% weight y-APS at pH= 2, (C) pH- 9, (D) pH -10.6, (E) pH -12.

Fig. 5 Calibration curve of partially cured y-APS by using area under CH2stretching bands.

Fig. 6 Calibration curve of silica wafer. The absorbance at 1970 cm-I

was plotted versus the weight in milligrams of silica.

Fig. 7 Number of y-APS molecules adsorbed on silica as a function of pH

of the treating solution. Highest adsorption occurs at pH- 10.6.

Fig. 8 FT-IR spectra of y-NPS (1% weight) at (A) pH= 2, (B) pH- 5,(C) pH= 12 treated on silica wafer.

Fig. 9 FT-IR spectra of vinyl silane (1% weight) at (A) pH- 2, (B) pH- 7.5,(C) pH- 12 treated on silica.

Fig. 10 Number of y-MPS molecules on silica treated from 1% weight solutionat various pH.

Fig. 11 Number of vinyl silane molecules on per milligram of silica treatedfrom 1% weight solution at various pH.

Fig. 12 Number of y-APS molecules deposited on E-glass as a function of pHof the treating solution (1% weight).

Fig. 13 FT-IR difference spectra: (A) silica treated with 1% weight y-APSat pH- 2, (B) pH- 10.6, (C) pH- 12 minus the untreated silica.Various y-APS structures, depending on pH of the treating solution,present on silica surface.

Fig. 14 FT-IR difference spectra of y-MPS treated on silica minus untreatedsilica. The band at 1705 cm-1 indicates hydrogen bonded carbonylgroup.

APSpH

12

10,8

9.9

9.6

9.3

8.9

7

1800 1400 1000 cm -Fig. 1

06

Cl, (

4C

0

kwa mIoad HO!S

AHmine dlef. SiOHpH(cm') (cm'1)

(CH 2 ) NH*ClI2

HO- Si -OH 11,101 1610 1505 902

OH

1008H

N H (HCO)Air dried HI-j-I-

iH 1575, 1488 930OH

N2 dried IC2)HO- Si-0-- H

I H 1601 9340H

Heated (CH2)r N H2

-Si-O-5i- 1591

H(CHg)BIN

O-Si-O 1595 925

0

Fig. 3

APS on Silica

Untreated

PH= 2

pH=9

PH 10. 6

3800 3000 2200 c m -

Fig. 4

C)

00

02

00

7i 0

0 0

(00

34

Bulqo44s ZHO &*punl DSJy

N0

0

2

SU

C0

00ha

A 'a

ac

1'ha

Og..a

U N

C',

I I a 0

4 N 0 0 * 4 N 0-

O~.6I &D *ouDqJosqy

am

0

00OCO)

C~Co

(023

K~~; D~IS~w /sSifoloSIwl

MPS on Silica

PH

5

12

3800 3000 2200 cm-,

* Fig. 8

Vinyl Silone an Silica

pHs 7.5

PH a 12

cm" 3800 3000 2200 1400

Fig. 9

* 06

0

CD

00

0 00

0

0 40

0 d

00

0*11S ;ogowsolno0*

NO*ON

00

0 w

0 (

C

0im3;

(/)lv 0 1*ol

0ad

0 E0

0 0

OD

0a0

00

4 Nn0t

- 0A

U (0

0. 0

CC

0I

*

00

InCl) 0a-'I-

o OLDE

0 0- .02

op~ *6*I

Cl) i34

C 000

00C

.0'em-0

00Cor")

-